space shuttle main engine start: what's the white plume at the edge of the engine bell?

Referencing the first ten seconds of this video of a space shuttle main engine start sequence, what are the white plumes drifting down from the edge of each engine bell?

Venting coolant or fuel, I believe.

The RS-25 Space Shuttle Main Engine (SSME) is a staged combustion engine using hydrogen fuel and liquid oxygen as oxidizer. As you can see from this diagram, the fuel is circulated through the nozzle to regenerate heat during engine operation, recovering thermal energy in the form of higher pressure on the hydrogen side necessary to get adequate flow to the high pressure turbopumps for hydrogen and LOX as well as kickstarting the low pressure fuel turbopump, so these appear to just be overpressure valves venting hydrogen. Note the “sparklers” firing from the sides which ignite the vented fuel to make sure it doesn’t disperse and form a detonable volume.

The white cones formed in the exhaust plume (which I don’t think is what the o.p. is referring to but often come into question) are shock diamonds (sometimes called Mach diamonds or disks) which are a result of interaction of oblique shock waves in overexpanded supersonic flow. It can also happen to a certain extent in underexpanded flow though it generally isn’t very visible. You sometimes see this portrayed in science fiction movies featuring a spacecraft flying through vacuum which is highly risible given that it is purely an atmospheric effect; without ambient pressure there is no free jet boundary. I suppose you could get something like it in a two phase flow where the lighter gases form sort of an envelop around the denser part of the plume that then expands or heats gas to supersonic speeds but it would have to be some really intense heating to get suffiicent velocity to get a shock condition inside of an unconstrained gas in vacuum. Basically, some GFX guy thought it looked cool and replicated the effect using CGI without understanding how it actually works. Anyway, Here is a pretty good explanation with some really nice diagrams someone went to the effort to draw up.


Of course, when you see venting anything, it is very likely to be condensed water vapor. Most gases, including hydrogen, are colorless. You see it because it’s very cold and ambient water vapor condenses into a cloud.

That is quite true; however, in the case of hydrogen, it is also producing aerosolized liquid nitrogen and oxygen from the air.


That’s surprising, but not shocking. I wouldn’t have expected there to be enough heat capacity in the cold hydrogen to do that. Though it takes a lot of heat transfer to condense water so I suppose it’s not a huge shock that it might be able to condense the air itself. The temperature difference might be higher for nitrogen but the enthalpy of vaporization is like 1/7 that of water.

When we were testing hydrogen fueled engines the last few feet of fuel line is corrugated flex line without insulation (to allow thrust measurements). These lines immediately begin raining oxygen condensed from the air. Sometimes they “poof” if they hit something. Not sure if any nitrogen was liquefied. After shutdown the lines were soon covered with several inches of water ice.


I can’t see where the di-lithium crystals go. :stuck_out_tongue:

Link goes to a 404 page.

And that’s a cartoonishly simplified diagram.

That said, the SSME is probably the most complex rocket engine ever developed. The equivalent diagram for a SpaceX Merlin 1D engine is here (that’s a generic “gas generator” engine, but they’re all basically the same in terms of propellant flow).

Well I didn’t see a hydrogen fire in the plume.
What you can do is spray the O2 around (to be sure there is heaps of O2 for the H2 to mix with )
What you can’t do is spray H2 around and then have a massive ball of pure (or below flammable O2) H2 drifting away from the ignition source.
Also, they may have used an inert gas to cool the system before putting H2 or O2 into it ? because rapid cooling may leak the H2 or O2 ? so are they putting inert ( N2 or He ?) through until pressure profile is tested as correct at operating temperatures ? - no damage, no leaks?

Usually Stranger is right on target, not sure this time though.

Pretty sure that the OP posted a video of oxygen being purged for igniting the system with hypergolic ignition.


From the above cite :

*They use a mix of two chemicals, triethylaluminum and triethylborane, aka TEA-TEB. Each is basically a metal atom (aluminum or boron) holding on to three hydrocarbon molecules (tri-ethyl), ready to break at a moment’s notice.

These two chemicals will spontaneously and near instantaneously burst into flame upon contact with oxygen. It can be oxygen in air or liquid oxygen in a rocket engine. The boron in the TEB is what causes the green flame when the engines start. To start the engine, LOX is flowed through the rocket injector into the chamber from the vehicle’s tank, TEA-TEB is injected into the chamber to create ignition, then RP-1 (fancy kerosene) is flowed in from the vehicle tank to start burning. The flows are increased, thrust is made, and the rocket launches.

To save weight on the first stage apparently they are flowing the TEA-TEB from a tank on the ground.*
I think I can see the TEA-TEB coming out of the side of the launch pad in the OP’s video.

I’m not sure what the point of your cite (which references a problem with the Merlin 1D engine on a SpaceX Falcon 9 vehicle) but the Space Shuttle Main Engines (SSME) do not use a TEA-TAB ignition or other pyrogenic system. The SSME, which uses LH[SUB]2[/SUB]/LOX propellant, has a capacitative pyrotechnic spark ignition system (Figures 7 thru 9).

As for the SSME ignition sequence, I’ll quote from Dennis R. Jenkins’ Space Shuttle: The History of the National Space Transportation System The First 100 Missions:

*At T-4 minutes, the system purge begins. It is followed at T-3 minutes 25 seconds by the beginning of the engine gimbal tests, during which each gimbal actuator is operated through a set profile of extensions and retractions. If all actuators function satisfactorily, the engines are gimbaled to a predefined position at T-2 minutes 15 seconds. The engines remain in this position until engine ignition.

At T-2 minutes 55 seconds the launch processing system closes the liquid oxygen tank vent vale, and the tank is pressurized to 21 psig with ground support equipment supplied-helium. The 21-psig pressure corresponds to a LO2 engine manifold pressure of 105 psia. At T-1 minute 57 seconds, the Launch Processing System closes the LH2 vent valve, and the tank is pressurized to 42 psig with ground-supplied helium. At T-31 seconds, the onboard redundant set launch sequencer is enabled by the launch processing system. From this point on, all sequencing is performed by the Orbiter GPCs in the redundant set. The GPCs still respond, however, to hold, resume count, and recycle commands from the Launch processing System.

At T-16 seconds, the GPCs begin to issue arming commands for the SRB ignition pyro initiator controllers, the hold-down release pyro initiator controllers, and the T-0 umbilical release pyro initiator controllers. At T-9.5 seconds the engine chill-down sequence is complete and the GPCs open the LH2 prevalves (the LO2 prevalves are open during loading to chill-down the engines).

At T-6.6 seconds the GPCs issue the engine start command and the main fuel and oxidizer valves in each engine opens. If all three SSMEs reach 90 percent of their rated thrust by T-3 seconds, then at T-0, the GPCs will issue the commands to fire the SRB ignition pyro initiator controllers, the hold-down release pyro initiator controllers, and the T-0 umbilical release pyro initiator controllers. If one of more of the three main engines do not reach 90 percent of their rated thrust by T-3 seconds, all SSMEs are shut down th SRBs are not ignited, and a pad abort condition exists.

I did find one amusing thing about your cite; the last sentence reads: “In theory, hypergols are more reliable. In practice, that seems to not necessarily be the case.” In fact, for those of use who have dealt with hypergolic pyrogenic igniters, the statement speaks to ignorance of the author or whomever he quoted. Hypergols are reliable in the sense that if you mix them together they will definitely ignite, but they require their own set of containers, valves, plumbing, et cetera, all of which can fail or be clogged. In comparison, a pyrotechnic igniter is generally either solid state device with no moving parts. The challenge with a dense fuel like kerosene or hydrazine is assuring sufficient mixing that the ignition cell expands rather than being extinguished in an oxidizer-poor environment.